EP1672595A1 - Procédé de détermination de la cause d'un état d'erreur sur un appareil - Google Patents

Procédé de détermination de la cause d'un état d'erreur sur un appareil Download PDF

Info

Publication number
EP1672595A1
EP1672595A1 EP05256484A EP05256484A EP1672595A1 EP 1672595 A1 EP1672595 A1 EP 1672595A1 EP 05256484 A EP05256484 A EP 05256484A EP 05256484 A EP05256484 A EP 05256484A EP 1672595 A1 EP1672595 A1 EP 1672595A1
Authority
EP
European Patent Office
Prior art keywords
detectors
program
atm
network
information
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
EP05256484A
Other languages
German (de)
English (en)
Inventor
Modupe Ayara
Simon James Forrest
Rogerio De Lemos
Jon Timmis
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
NCR International Inc
Original Assignee
NCR International Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by NCR International Inc filed Critical NCR International Inc
Publication of EP1672595A1 publication Critical patent/EP1672595A1/fr
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07F—COIN-FREED OR LIKE APPARATUS
    • G07F9/00—Details other than those peculiar to special kinds or types of apparatus
    • G07F9/02—Devices for alarm or indication, e.g. when empty; Advertising arrangements in coin-freed apparatus
    • G07F9/026—Devices for alarm or indication, e.g. when empty; Advertising arrangements in coin-freed apparatus for alarm, monitoring and auditing in vending machines or means for indication, e.g. when empty
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
    • G07D11/00—Devices accepting coins; Devices accepting, dispensing, sorting or counting valuable papers
    • G07D11/20—Controlling or monitoring the operation of devices; Data handling
    • G07D11/22—Means for sensing or detection
    • G07D11/235—Means for sensing or detection for monitoring or indicating operating conditions; for detecting malfunctions
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07F—COIN-FREED OR LIKE APPARATUS
    • G07F19/00—Complete banking systems; Coded card-freed arrangements adapted for dispensing or receiving monies or the like and posting such transactions to existing accounts, e.g. automatic teller machines
    • G07F19/20—Automatic teller machines [ATMs]
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07F—COIN-FREED OR LIKE APPARATUS
    • G07F19/00—Complete banking systems; Coded card-freed arrangements adapted for dispensing or receiving monies or the like and posting such transactions to existing accounts, e.g. automatic teller machines
    • G07F19/20—Automatic teller machines [ATMs]
    • G07F19/207—Surveillance aspects at ATMs

Definitions

  • the invention relates to a method of determining the cause of an error state in an apparatus, and has particular application, for example, to use in determination of errors in self service terminals (SST) such as automated teller machines (ATM).
  • SST self service terminals
  • ATM automated teller machines
  • the invention has particular application to the analysis of causes of error states in an ATM, for the sake of clarity, the invention will be described with reference to an ATM and to a network of ATMs. However, the invention can be applied to the operation of any apparatus or device as well as any network of such apparatuses or devices.
  • a standard ATM having the facility to dispense bank notes includes electronic control means connected to both a currency dispenser unit and a user interface device.
  • a user inserts a user identity card into the machine and then enters certain data, such as a personal identification number (PIN) and the quantity of currency required to be dispensed, by means of a key pad incorporated in the user interface device.
  • PIN personal identification number
  • the ATM will then process the requested transaction, dispense notes extracted from one or more storage cassettes within the currency dispenser unit, update the user's account to reflect the transaction and return the card to the user as part of a routine operation.
  • various malfunctions may occur from time to time. For example, bank notes may become jammed in the feed path, the pick means, utilised to select a note from an ATM currency cassette, may fail to pick a bank note from the associated storage cassette, or there may occur multiple feeding in which two or more notes are fed in superposed relationship to the stacking means.
  • the problems discussed above may be caused by wear of components in the dispenser unit or by changes in the ambient conditions in the vicinity of the ATM.
  • the ATM may be shut down until the malfunction is rectified, which will require the intervention of a trained operator, or in the event of multiple feeding the picked notes will be diverted to a purge bin resulting in less efficient operation of the ATM.
  • a method of determining the cause of an error state for one or more components within an apparatus comprising a plurality of sensors arranged to monitor the operation of components of the apparatus and a control means arranged to receive said information from said plurality of sensors, the method comprising: a) analysing said sensor information in the form of an error log to ascertain sensor patterns from said sensor information; b) comparing said sensor patterns with detectors, which are predefined patterns, indicative of the condition of said one or more components within the apparatus; and c) classifying said sensor patterns as being indicative of said error state of a component or not based upon a comparison of sensor patterns with said detectors.
  • a computer program for determining the cause of an error state of one or more components within an apparatus comprising a plurality of sensors arranged to monitor the operation of components of the apparatus and a control means arranged to receive said information from said plurality of sensors, the program being adapted to: a) analyse said sensor information in the form of an error log to ascertain sensor patterns from said sensor information ; b) compare said sensor patterns with detectors, which are predefined patterns, indicative of the condition of said one or more components within the apparatus; and c) classify said sensor patterns as being indicative of said error state of a component or not based upon a comparison of sensor patterns with said detectors.
  • the solution provides a system for automatically extracting sequences of states that will lead to a fatal state from the log file produced by the modules in an ATM.
  • the system incorporates a learning capability and a set of databases that can store learning applicable to specific models of an ATM, in a particular environment, or ATMs with a particular usage pattern.
  • the system also provides an automated learning capability that allows the system to detect novel error sequences and thus improve the accuracy of the error state detection. It also provides a library of databases that can be used to analyse the log files from different models or families of ATM.
  • the system can incorporate detectors and rules for specific environmental conditions, such as cold climates.
  • the system provides graphical methods of displaying the information in the log file at a high level to allow an overview of the ATM behaviour.
  • the system is compatible with all ATM log files and can incorporate information from other log files to provide a higher level view of the ATM behaviour prior to failure.
  • an ATM 2 which includes a control means in the form of a central processor unit (CPU) 4 which has stored therein a control program which controls the operation of the ATM 2 in dependence upon information gained from a plurality of sensors 110-120. If sensors are added or removed from the terminal 2 the program may be updated. The program monitors and optimises the operation of the ATM 2.
  • CPU central processor unit
  • the CPU 4 is connected to a user interface device 6 incorporating a slot 8 (FIG. 1), connected to a conventional card reader 130 (FIG. 4), for receiving a user identity card, a key pad 10 for inputting data, a screen 12 for displaying user information, and an output slot 14 for dispensing bank notes to a user.
  • the CPU 4 is also connected to a cash dispenser unit 16 (FIG. 2) and a conventional printer 122 (FIG. 4) for printing documents such as statements, receipts and account balances.
  • the cash dispenser unit 16 includes two similar pick means 18 arranged one above the other and respectively associated with two storage cassettes 20 which are removably mounted in a supporting framework 22 of the dispenser unit 16.
  • Each of the storage cassettes 20 is arranged to contain a stack of bank notes 24, corresponding long edges of which are supported on a horizontal support plate 26 mounted in the storage cassette 20.
  • the stack of notes 24 in each storage cassette 20 is urged by a spring loaded pusher member 28 towards a stop member 30 mounted at the front end of each storage cassette 20.
  • An opening 32 is formed in the front end of each storage cassette 20, the opening 32 being closed normally by conventional shutter means (not shown) when the storage cassette 20 is not mounted in the dispenser unit 16.
  • the shutter is automatically retracted to enable notes 24 to be extracted through the opening 32 by the associated pick means 18.
  • Each pick means 18 includes a tubular member 34 which extends between, and is rotatably mounted with respect to, side walls 36 and 38 (FIG. 3) of the framework 22.
  • Two conventional pick arms 40 are secured on each tubular member 34, each pick arm 40 communicating with the interior of the associated tubular member 34.
  • Corresponding ends of the tubular members 34 project beyond the side wall 38, and are each connected by a respective swivel elbow connector 44 to a respective rubber tube 46 via which reduced pressure is applied in operation to the respective tubular member 34.
  • the suction force produced by the suction pump 140 (FIG.
  • a gear segment 48 is secured to that part of each tubular member 34 projecting beyond the side wall 38, the gear segment 48 being in co-operative engagement with a toothed end portion 50 of a first arm of a respective bell crank lever 52 which is pivotably mounted on a stud 54 secured to the outer surface of the wall 38.
  • Each lever 52 is urged to rotate in a counter clockwise direction with reference to FIG. 3 by means of a spring 56 the ends of which are respectively attached to the side wall 38 and to the end of the second arm of the lever 52.
  • a stud 58 is secured to one side of each lever 52, the stud 58 engaging in a cam track 60 formed in an associated cam member 62.
  • Each cam member 62 is secured to a respective gear wheel 64 which is rotatably mounted on a respective shaft 66 projecting from the outer surface of the side wall 38.
  • the gear wheels 64 are driven by gear wheels 68 forming part of a gear mechanism 69 operated by a main electric drive motor 70 (FIG. 4).
  • the gear wheels 64 In operation (with the drive motor 70 energised) the gear wheels 64 are rotated in a clockwise direction with reference to FIG. 3.
  • This rotation of the gear wheels 64 brings about an oscillatory pivotal movement of the levers 52 by virtue of the engagement of the studs 58 in the cam tracks 60, the springs 56 holding the studs 58 in engagement with the inner edges of the cam tracks 60.
  • the oscillatory movement of the levers 52 brings about an oscillatory pivotal movement of the assemblies of the tubular members 34 and the associated pick arms 40.
  • the oscillatory movement of either of the assemblies of the tubular members 34 and the associated pick arms 40 is effective to cause notes 24 to be picked one by one from the stack of notes 24 held in the associated storage cassette 20.
  • the ATM 2 incorporates a motor sensor 110 which includes a timing disc 72 (FIG. 3) secured to the face of each gear wheel 60 remote from an associated cam member 62.
  • the timing disc 72 is for the most part transparent but incorporates an arcuate opaque strip 74 extending around just over half the periphery of the disc 72.
  • Each timing disc 72 is associated with optical sensing means, comprising an LED (not shown) and a co-operating photo-transistor sensor 112, which is arranged to sense the opaque strip 74.
  • the associated sensor 112 In operation, as each assembly of a gear wheel 64 and the associated cam member 62 and timing disc 72 rotates in response to energizing of the drive motor 70, the associated sensor 112 generates output signals in response to the sensing of the leading and trailing edges of the associated opaque strip 74. It should be understood that the signals generated by each of the sensors 112 provide indications as to the precise positions of the associated pick arms 40 at the times when these signals are generated.
  • the speed of rotation of the drive motor 70 can be varied in order to vary the time for which the pick arms 40 hold the associated suction pads 42 in contact with a first note 24' in the stack of notes 24 in one of the storage cassettes 20, before attempting to pick the first note 24' from the storage cassette 20. If the solenoid valve 142 is opened just after the suction pads 42 are brought into contact with the first note 24' then varying the period for which the suction pads 42 are held in contact with the first note 24' will vary the suction force applied to the first note 24', as will be discussed in more detail below.
  • the suction force applied to the first note 24' prior to attempting to pick the first note 24' from the storage cassette 20 can also be varied by varying the delay prior to opening the solenoid valve 142 to apply the suction force to the first note 24'.
  • the suction pump 140 (FIG. 4) operates continuously the longer the delay prior to opening the solenoid valve 142 the larger the suction force produced by the suction pump 140 will be.
  • the suction force used in picking the first note 24' can be varied by varying either the speed of rotation of the drive motor 70 or varying the delay prior to opening the solenoid valve 142.
  • the dispenser unit 16 also incorporates feed rollers 77 for feeding the bank notes 24 along a feed path 78 from each of the storage cassettes 20 to a stacking wheel 82 and on to the output slot 14, the rollers 77 being associated with co-operating first and second rollers 79 and 80 which are positioned at the opening 32 in the front of each storage cassette 20.
  • the stacking wheel 82 is arranged to receive notes 24 fed along the feed path 78.
  • the stacking wheel 82 serves to stack notes 24 picked from one or both of the storage cassettes 20 so as to form a bundle 84 of notes for delivery to the output slot 14 for collection by the user.
  • the stacking wheel 82 is driven by the drive motor 70 and is arranged to rotate continuously in operation in a counter clockwise direction. Means (not shown) are provided between the upper transport mechanism 85 and the stacking wheel 82 for detecting any multiple feeding of notes and for detecting any invalid or torn note.
  • the stacking plates 86 are spaced apart in parallel relationship along the stacker wheel shaft 88, each stacking plate 86 incorporating a series of curved tines 90. The tines 90 of the stacking plates 86 pass between portions of a rockably mounted stripper plate assembly 94.
  • each note fed along the feed path 78 to the stacking wheel 82 enters between adjacent tines 90 and is carried partly around the axis of the stacking wheel 82, the note being stripped from the wheel 82 by the portions of the stripper plate assembly 94 and being stacked against belt means 95.
  • the belt means 95 co-operates with belt means 98 normally held in the position shown in FIG. 2.
  • the belt means 98 is rocked in a clockwise direction about a shaft 100 so as to trap the bundle 84 of notes between the belt means 95 and the belt means 98. It should be understood that in the course of this rocking movement separate belts making up the belt means 98 pass between adjacent pairs of the stacking plates 86.
  • the belt means 95 and 98 are operated so as to drive the bundle 84 to an adjacent pair of belt means 102 and 104.
  • the belt means 102 and 104 serve to drive the bundle 84 through the output slot 14 to a position where the bundle 84 can be collected by the user of the ATM 2, a shutter 106, which serves to close the slot 14 when the ATM is not in operation, having previously been retracted to an open position.
  • the belt means 95 and 98 are mounted in resilient relationship relative to each other, and the belt means 102 and 104 are also mounted in resilient relationship relative to each other, so that bundles of notes of varying thickness can be held between, and fed by, the belt means 95 and 98 and the belt means 102 and 104.
  • the belt means 95, 98,102 and 104 are driven under the control of the CPU 4 by a bi-directional stepping motor 71.
  • the stripper plate assembly 94 is rocked into the position shown in chain outline in FIG. 2, and the belt means 95 and 98 are operated to feed the bundle 84 in a direction opposite to the normal feed direction, the bundle 84 being deposited in a purge bin 108 via an opening in the top thereof.
  • the stepping motor 71 can be operated so as to cause the belt means 95, 98,102 and 104 to drive the note 24 or bundle 84 of notes in the forward and the reverse direction repeatedly, in an attempt to unblock the currency jam or to realign the bank note 24 or bundle 84 of bank notes.
  • An ATM 2 in accordance with the present invention incorporates a plurality of sensors 110-120 (FIG. 4) in communication with the CPU 4 arranged to monitor the operation of the ATM 2 and the ambient conditions.
  • the CPU 4 is adapted to alter the operation of the ATM 2 in dependence on the output of the sensors 110-120 so as to reduce the number of malfunctions that occur in operation.
  • the sensors 110-120 comprise: a first motor sensor 110 located adjacent the drive motor 70 and a second motor sensor 112 located adjacent the stepping motor 71, the first motor sensor 110 including a photo-transistor sensor 113 (FIG.
  • the second motor sensor 112 including a photo-transistor sensor (not shown) arranged to detect the speed and rotational direction of the stepping motor 71; a purge bin sensor 114 located adjacent the entrance to the purge bin 108 and arranged to detect the deposition of a single note 24 or a bundle 84 of notes in the purge bin 108; a plurality of optical bank note location sensors 116 located along the feed path 78 and between the stacking wheel 82 and the output slot 14 and arranged to monitor at any instant the presence or absence of notes 24 at different locations within the ATM 2; a plurality of temperature sensors 118 located within the ATM 2, providing the CPU 4 with an accurate measure of the temperatures at selected locations throughout the ATM 2; and a plurality of humidity sensors 120 also located within the ATM 2 so as to provide the CPU 4 with an accurate measure of the ambient humidity at selected locations throughout the ATM 2.
  • the sensors 110-120 When the ATM 2 is operating, the sensors 110-120 continually monitor the operation of the ATM 2 and ambient conditions and communicate the information obtained to the CPU 4. For example, the temperature sensors 118 may detect that the ambient temperature within the ATM 2 is lower than a predetermined temperature. On receipt of this information the CPU 4 will bring about one or more of a number of actions in order to reduce the likelihood of a malfunction occurring. Thus, for example the CPU 4 may reduce the speed of the drive motor 70 which drives the rollers 77, 79, 80 thereby reducing the likelihood of slippage between a note 24 and the rollers 77, 79, 80 while the note 24 is being fed through the dispenser unit 16.
  • the drive motor 70 also controls the positioning of the pick arms 40, reducing the speed of the drive motor 70 will cause the rubber suction pad 42 of the pick arms 40 to be held adjacent the first note 24' in the corresponding storage cassette 20 for an increased period of time thereby increasing the suction force applied to the note 24'.
  • the exact increase in time that the rubber suction pads 42 are held in contact with the first note 24' prior to picking will depend on the ambient temperature detected by the temperature sensors 118.
  • the time that suction is applied by the suction pads 42 to the first note 24' is accurately monitored by the CPU 4 through the photo-transistor sensor 112, which detect the speed of rotation of the motor 70 and consequently the location of the pick arms 40 and the associated suction pads 42.
  • the CPU 4 may increase the suction force applied to the first note 24' by increasing the delay prior to opening the solenoid valve 142 to apply the suction force to the first note 24', as discussed above.
  • the CPU 4 obtains temperature information from each of the temperature sensors 118 which can be processed separately so that the CPU 4 can vary the operation of individual components of the ATM 2 dependent on their temperatures so as to optimize the operation of the ATM 2.
  • a temperature sensor 118 is located in each of the storage cassettes 20 and at various locations throughout the feed path 78. If the first storage cassette 20 is at a higher temperature than the second storage cassette 20 a note 24 will be picked from the second storage cassette 20 more slowly than from the first storage cassette 20 in order to compensate for the lower temperature in the second storage cassette 20.
  • the feed means 77 can be controlled differently in different sections of the feed path 78 in order to compensate for differences in ambient temperature detected by the temperature sensors 118 located throughout the feed means 78.
  • the CPU 4 also monitors by means of the sensor 114 the deposition of a note 24 or a bundle 84 of notes in the purge bin 108. If the CPU 4 finds that the rejection rate is tending to increase then the CPU 4 will cause the speed of the drive motor 70 to be reduced, which action will normally be successful in reducing the rejection rate. Under the control of the control program stored therein, the CPU 4 maintains the time taken to dispense a bundle 84 of notes as low as possible while limiting the number of times that notes 24 are rejected to a predetermined acceptable percentage of total pick operations.
  • a feature of the ATM 2 when operated in accordance with prior art operational methods is that the operating characteristics and ambient conditions of the ATM 2 are monitored at given times and its operation is altered in dependence thereon in order to optimise its operation at that time.
  • detectors during the operation of an apparatus, such as an ATM
  • an apparatus such as an ATM
  • detectors can be thought of as an architecture and implementation of an Artificial Immune System (AIS) to provide an Adaptable Error Prediction System (AEPS) that will add intelligence, learning and predictive capabilities to the processing of device status information provided by the modules in an ATM.
  • AIS Artificial Immune System
  • AEPS Adaptable Error Prediction System
  • the architecture is distributed throughout the network with agents on the individual ATMs in the network and a central management system that co-ordinates the processing of the information reported from the agents (Fig. 6).
  • Each ATM has a local AEPS implemented as an AIS for local monitoring of the device data. These send their data through the ATM network to a network-wide AEPS which is implemented as a central AIS in the network management system.
  • AIS AIS
  • the framework also divides the learning mechanisms into the two levels of: (1) learning within an ATM through the local AEPS and (2) learning amongst ATMs through the network-wide AEPS.
  • the design-time immunisation phase caters for the distribution of generic detectors amongst ATMs, from an off-line process of detector generation.
  • the run-time adaptation phase confers on each ATM a more specialised set of detectors and is responsible for augmenting the generic detectors.
  • the detectors in this case, are pattern recognisers that are endowed with the capabilities for detecting patterns in the ATM device data.
  • An ATM in the network would initially be provided with a set of generic detectors, hence the term immunisation.
  • the generic detectors are then augmented with new information in the run-time adaptation phase, such that an ATM is conferred with the ability to learn new patterns. This is based on the definition of learning, which is defined as the augmentation of existing information with novel information.
  • An overview of the system is illustrated in Figure 5.
  • the off-line process for generating detectors can use historical data based on patterns detected in a current ATM where historical log data (Table 1) is available. They can also be generated during the development of a new ATM. This allows design engineers to optimise the detectors that are used to seed the ATM during the design process. This optimisation is based on the engineers experience and knowledge of the modules and the state transitions that can generate error conditions. It allows them to remove or tune the detectors generated to provide the optimum set of detectors.
  • the detector generation process can be simply described as learning from the past trends in the system to make inferences on when future states of the system may lead to a fatal state. The outcome of this process is a set of generic detectors that are capable of detecting fatal errors common to two or more models of ATM. Immunisation is the process of injecting into the local AEPS of the ATM the detectors that were generated off-line and is aimed at distributing generic detectors to all the local AEPSs in the ATMs.
  • the first part is the error detection, where the detectors monitor error behaviour in an ATM.
  • the state information generated by the real-time behaviours of the ATM is passed to the local AEPS for classification.
  • This process is performed by classifying incoming states from the device data into states that will induce a fatal state and those that will not.
  • the process of classifying the states is based on a comparison of sequences of incoming states with the existing detectors for a match. This may give rise to situations when the current state in the sequence of incoming states results in no matching detectors. This can then be classified as a novel sequence. Alternatively, the current state may give rise to a new sequence with a matching detector.
  • confidence values are associated with the detectors. These confidence values influence the decision for selecting detectors such that the detector with the highest value is selected. Adaptation of these confidence values can be performed with regard to correct or incorrect inferences by the associated detectors. This implies that a detector that provides correct inferences from classifying a sequence is rewarded, but penalised for incorrect inferences.
  • the classification of the ATMs error behaviours may induce appropriate actions if the behaviours are inferred to be precursors to a fatal state. These actions are determined by the expected time that the fatal event will occur. Therefore, actions are initiated when the time interval between the detection of error behaviour and expected time of occurrence of the fatal event lies between a defined significant time interval (e.g. minimum of an hour). Thus, the defined time interval signifies the minimum time within which an alert can be triggered. Alternatively, a fail-safe method could be applied shutting down the system to prevent damage.
  • the minimum information contained in the alert should be the inferred fatal state of the ATM as well as expected time of occurrence, which is evaluated for authenticity. This evaluation of the alerts could be through the application of information from system maintenance status or by human experts.
  • the detectors learns the new error behaviour with a view to generating representative detectors.
  • the learning process is achieved through continuous rule mining. This is an on-line rule generation algorithm that can be applied to generating new rules representing novel patterns. Depending on the representation of the detectors other algorithms could be applied to generate the representation of the novel patterns.
  • the outcome of the learning process is a set of new detectors, called immature detectors that are initially subject to local tolerisation and then local validation before being incorporated into the local AEPS.
  • the local tolerisation process takes the representation produced from the learning process and selects immature detectors that are competent enough to be incorporated into the local AEPS. It is based on the criterion of proving competency at correctly classifying patterns. This is performed by evaluating if a new detector correctly classifies a pattern as leading to a fatal or nonfatal state thereby leading to its incorporation into the local AEPS, otherwise it is discarded. This process occurs within a stipulated period (lifespan) within which the immature detector is expected to prove its competency. At this stage in the processing of the detectors a copy of the new tolerised detector is propagated to the network-wide AEPS.
  • an immature detector survives local tolerisation then it is locally validated to confirm its meaningfulness. This is achieved by taking the immature detectors that detect erroneous behaviour and validating them to ascertain the accuracy of the detection. This is carried out by a human-expert in the related domain either a field engineer or the detectors could be validated by a subject matter expert for a specific ATM module. This is used as a method of providing feedback on novel error states to NCR engineering from ATMs in the field. The validation could also be carried out within the local AEPS using automated methods that can be applied to validate the detectors. This allows the testing of the new detector against existing detectors to detect conflicts or contradictions.
  • the automated system can also validate the new detector against a set of known healthy states stored in the local AEPS to ensure that the new detector does not misclassify these as error states. Due to the complexity of including a human in the real-time processing of the new detectors, the best solution may be an automated validation process carried out in the local AEPS based on stored domain-knowledge and a set of business rules.
  • the new detectors can also be validated off-line by either an NCR field engineer, WCS or NCR Engineering as an additional means of filtering invalid detectors.
  • the competent detectors have survived local tolerisation and local validation they are incorporated into the local AEPS.
  • they are added to groups of similar detectors, where similarity is based on defined criteria.
  • the flow of the detectors is illustrated in Figure 5.
  • the incorporation of the new detectors into relevant groupings is through the application of a clustering algorithm.
  • An example of this is to apply a nature-inspired learning algorithm - Self-Stabilising Artificial Immune Systems (SSAIS), or meta-stable memory algorithm for incorporating the new detectors.
  • SSAIS Self-Stabilising Artificial Immune Systems
  • These algorithms are AISs that are able to maintain populated regions of the detectors as clusters. Subsequently, a copy of the new detector is also propagated to the network-wide AEPS.
  • the process is evaluated by calculating statistical data such as classification accuracy, population of generic detectors, population of specialised detectors, proportion of classification accuracy accounted for by population of generic detectors versus specialised detectors, and true positive versus false positive detection ratio. These calculated values are also propagated to the network-wide AEPS for global evaluation of local AEPSs. This is a background process to provide information on the detection performance of the detectors in the local AEPS.
  • the evaluation of the local AEPS in the on-line network process forms part of the criteria for evaluating detectors in both the local and network-wide AEPS.
  • Alerts triggered by the new detectors (immature detectors) in the local AEPS systems are forwarded to the network-wide AED for evaluation of their authenticity.
  • Immature-network detectors, in the network AEPS are also evaluated for their authenticity within this process.
  • the evaluation in the network-wide AEPS is carried out in a similar manner to the error detection in the local AEPS but with a network perspective. This allows the comparison between inputs from the different local AEPSs as well as evaluating the inputs from each local AEPS.
  • the evaluation process results in the rewarding or penalisation of the detectors in local AEPSs that triggered the alerts. Detectors in local AEPSs that have initiated an alert based on the classification of a state as fatal are rewarded for correct alerts, while incorrect alerts are penalised.
  • the network tolerisation of immature-network detectors in the network-wide AEPS occurs within a specified period during which immature-network detectors have to display their competency at the network level.
  • the process of network tolerisation for each immature-network detector involves a count of local AEPSs that have propagated similar immature-network detectors.
  • the copy of the immature-network detector in a local AEPS is expected to have correctly classified error patterns in the local AEPS.
  • the network validation is similar to the local validation since they both involve feedback from an expert, but in this case feedback is provided on the immature-network detectors that have survived network tolerisation.
  • the feedback determines whether such immature-network detectors will be promoted to competent-network detectors or not. The outcome is that immature-network detectors that have successfully undergone network tolerisation and network validation become competent-network detectors.
  • the competent-network detectors generated are applied to immunise the local AEPSs.
  • the process only applies to the local AEPSs that do not currently have a copy of the competent-network detectors. These will be the ATMs that have until that point not detected the pattern of state transitions that can lead to the fatal state in the ATM.
  • the immunisation process extracts generic detectors from the pool of new detectors for distribution to all local AEPSs in the network. It serves as a means of updating the generic detectors in all the local AEPSs.
  • the processing stages that apply to the device level ATM data and were described previously require a communication mechanism within the ATM network to allow the communication of the detectors generated in the local AEPS to be passed to the network-wide AEPS and the network-wide AEPS to immunise the various local AEPSs in the ATM network.
  • This communication can be supported by the management infrastructure currently used for the management of ATM networks.
  • the framework used to implement the network based AIS system does not require any changes to the current architecture and can be used in parallel with the existing management systems. An overview of the architecture is shown in Figure 6.
  • Each ATM on the network contains a local AEPS implemented as a software AIS agent.
  • the AIS monitors the behaviour of the ATM through error state information, known in the field as the M-Status and M-Data, which is reported from each device.
  • M-Status and M-Data error state information
  • M-Data error state information
  • This data is already processed in the application since it is used to generate an error log, known in the field as a devlog file, so the implementation of the AIS only requires the data to be passed to it as well.
  • Other sources of data from other standard log files that are written during a transaction can also be used to augment the device data.
  • the AIS agent contains the local copies of the detectors and the immature detectors that are undergoing tolerisation.
  • the agent When the agent detects an error condition the alert that is generated will be sent through the existing management interface.
  • This could be the Simple Network Management Protocol (SNMP) interface connected to a NCR Gasper TM management system.
  • SNMP Simple Network Management Protocol
  • the alert would be wrapped in the existing management protocol and sent to the management centre in a similar manner as any other alert.
  • the central ATM management system contains the network-wide AEPS, again implemented as an AIS as described above. This can co-exist with the existing management and dispatch system used by the financial institution. It can use the ATM network SNMP manager to receive and send query messages to the AIS agents in the ATMs in the network.
  • the AIS network monitors the performance of the immature network detectors in its tolerisation area. When these have been tolerised they are then propagated through the network to all ATMs. These can be applied to all the ATMs of a specific model, in similar usage patterns or to all ATMs on the network. The information can also be sent to an engineer to be applied to either new ATMs or to be propagated to other network-wide AEPSs. In this way the learning from one system can be used across all AIS enabled systems and can be included in the generic detectors that are used to seed new ATMs. The propagation of the new detectors would also be through the existing management system again using SNMP if this was available. This allows the AIS to be integrated with an existing ATM network without requiring additional infrastructure to be implemented.
  • the architecture also addresses the current problem of "false positives" that can be generated from the state information.
  • the local AEPS creates a warning based on a prediction of a potential fatal state, this will then be processed by the network AEPS prior to being passed onto the network management software. This allows these alarms to be filtered and the predictions of the local AEPS to be tuned by the network AEPS through the application of intelligence that has been gathered from all the local AEPSs on the network.
  • the network AEPS can also be used to tune the local AEPSs in the timing of predicted fatal states. This will again take input from a number of local AEPSs to intelligently process the timing of the predictions of an ATM entering a fatal state. This will be very important to allow timely preventative maintenance on ATMs that are predicted to fail. The exact prediction of the time to failure will allow the scheduling of the service call to avoid unnecessary dispatches of field engineers to the ATM. This timing information will be built into the system as well as the prediction based on the next state to predict when the state will change.
  • the invention provides an off-line log file analysis tool that can be used by engineers to more efficiently and successfully determine the sensor state changes that lead to particular failure modes or error states.
  • the system is based on an Artificial Immune System (AIS) which is a novel biological inspired software programming metaphor that allows intelligence to be built into the software system.
  • AIS Artificial Immune System
  • the application of this metaphor to the analysis of ATM log files allows the development of a system that incorporates pattern matching for automatically detecting state sequences and provides the ability to dynamically learn from the data passed through the system. This ability improves the pattern matching capability incrementally since each new log file analysed adds to the intelligence in the analysis system.
  • the system can also automatically build a set of databases based on the ATM model, the environment that the ATM operates in or the specific usage pattern for an ATM. This provides the system with the capability to tailor the analysis for a specific log file depending on the ATM model family or where in the world the ATM is situated taking into account the external environmental conditions. In this way different sets of rules can be applied to the log file data incorporating learning from other ATMs which have developed problems in similar weather conditions, due to its environment
  • the AIS system used for the off-line log analysis uses an Adaptive Error Detection System (AEDS) to generate a set of detectors based on patterns of states in the log data that lead the ATM into a fatal state that would take it out-of-service.
  • AEDS Adaptive Error Detection System
  • the system is provided at module and ATM design time with an initial set of detectors for patterns of errors that are defined as fatal states or are known to lead to fatal states.
  • AEDS Adaptive Error Detection System
  • the feedback from this testing allows additional data to be added to the system to recognise other patterns leading to fatal states found during the testing phase.
  • These initial detectors are stored in the database relevant to the ATM model or ATM family which can then be used to detect patterns that lead to fatal states.
  • These databases can then be used when a new log file is processed by the log analysis system. When a new log file is analysed the patterns of states leading to a fatal state are inferred from the provided log data.
  • the system also has the capability to learn through the detection of novel patterns that cause fatal states. When a new pattern is discovered a new detector is generated by the system for this error condition. This is then held for tolerisation, which is a process for identifying and confirming the meaningfulness of the new detector, which is referred to as an immature detector.
  • the detectors generated during the initial design and testing of a new module or ATM and also during the actual analysis of the log files can be represented as a set of rules. These rules could be generated by the application of a learning algorithm such as the Continuous Association Rule Mining Algorithm (CARMA) or some other algorithm.
  • CARMA Continuous Association Rule Mining Algorithm
  • the detectors generated as rules are then classified and clustered after tolerisation using a learning algorithm such as a Self-Stabilising Artificial Immune System (SSAIS).
  • SSAIS Self-Stabilising Artificial Immune System
  • Annex A is an example of a dev log file.
  • An analysis of the log file in accordance with the present invention is carried out, as discussed above, in order to determine a repeating pattern of states that had been recovered from automatically by the ATM that eventually lead to a fatal state. For example, if the system detected a plurality of M-Status 8 errors, meaning the purge bin overfill sensor was blocked followed by a fatal M-Status of 10 meaning too many errors or a M-Status 18 meaning a currency jam in the presenter, either of these can be caused by the previous recovered from M-Status 8's.
  • the detectors are not hard coded into the software, if a new sensor or module is included, the detectors for this can easily be incorporated into the databases. This allows the databases to be extended without having to re-create any of the log analysis code.
  • the log analysis system is also backwardly compatible with all the existing ATMs since it is an off-line process and does not require a specific ATM platform to run.
  • the system also incorporates a visualisation tool that can be used to display the clustering of rules that fulfil similar criteria. These will be defined by the NCR engineers and also created automatically by the learning algorithm. This can be used to show the clustering of the sequences of states in the log file providing a high level view of the data within the file. It also has the ability to provide frequency analysis of the sequences of state changes providing a view of the number of occurrences of the various states within a fixed time period. Other data can also be extracted from the log file and displayed graphically providing the engineer with high level information on the contents of the log file. See Figure 8. This can be integrated with translation software that would allow the original encoding from the log file to be displayed in a more human readable and easier to read format.
  • This information can provide an overview of the behaviour of the ATM and highlights recurring problems as a series of error states, even if the states have been recovered from and not caused an fatal error. This provides a "first glance" of the recent behaviour of the SST without having to read through the actual log file entries.
  • the information displayed can also be incorporated with other streams of data from other logs. This could provide a view of what transaction was being carried out, for example, allowing a view of the state of the whole ATM during a particular device problem. This allows cross-verification of problem causes and may highlight the root cause of a problem rather than trying to extract it from the actual device error log data.
  • This system can use the time stamps in the log data to provide additional information on the sequences of states that lead to the final fatal state. This allows the system to predict sets of state changes that can generate error states and these can be extrapolated from the log file during the detection of novel state sequences. This information can then be passed to engineers to be included in the hard coded error conditions that are used for ATM and module control. The system can also highlight when a number of state sequences were detected before the final sequence that lead to the fatal state. This information can be important in the diagnosis of a recurring problem in an ATM. The number of sequences from a number of different log files from different ATMs, within a network of ATMs, can also be extracted and collated by this system which again is difficult within the current manual system.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
EP05256484A 2004-12-17 2005-10-19 Procédé de détermination de la cause d'un état d'erreur sur un appareil Withdrawn EP1672595A1 (fr)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
GBGB0427694.5A GB0427694D0 (en) 2004-12-17 2004-12-17 A method of determining the cause of an error state in an apparatus

Publications (1)

Publication Number Publication Date
EP1672595A1 true EP1672595A1 (fr) 2006-06-21

Family

ID=34090246

Family Applications (1)

Application Number Title Priority Date Filing Date
EP05256484A Withdrawn EP1672595A1 (fr) 2004-12-17 2005-10-19 Procédé de détermination de la cause d'un état d'erreur sur un appareil

Country Status (3)

Country Link
US (1) US7600671B2 (fr)
EP (1) EP1672595A1 (fr)
GB (1) GB0427694D0 (fr)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2088564A1 (fr) * 2008-01-31 2009-08-12 NCR Corporation Terminal en libre-service
DE102009032023A1 (de) * 2009-07-07 2011-01-13 Giesecke & Devrient Gmbh Banknotenbearbeitungsmaschine mit einer Kassette
EP2342698A2 (fr) * 2008-09-11 2011-07-13 Non Linear Concepts, Inc. Diagnostic de système intelligent pour machines de traitement de billets de banque
US8078912B2 (en) 2008-01-31 2011-12-13 Ncr Corporation Self-service terminal
WO2014186472A1 (fr) * 2013-05-14 2014-11-20 Diebold Self-Service Systems Division Of Diebold, Incorporated Guichet automatique qui peut fonctionner pour détecter et stocker automatiquement des activités de services
KR20210063397A (ko) * 2018-11-27 2021-06-01 후지츠 프론테크 가부시키가이샤 원인 추정 장치, 원인 추정 출력 방법 및 지엽류 취급 시스템

Families Citing this family (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8157078B1 (en) * 2008-11-25 2012-04-17 Bank Of America Corporation Cash handling device having environmental condition monitoring system
US8738973B1 (en) 2009-04-30 2014-05-27 Bank Of America Corporation Analysis of self-service terminal operational data
US8554714B2 (en) * 2009-05-11 2013-10-08 Honeywell International Inc. High volume alarm management system
DE102009021130A1 (de) * 2009-05-14 2010-11-18 Wincor Nixdorf International Gmbh Einrichtung zum zentralen Überwachen des Betriebes von Bankautomaten
US7926809B2 (en) * 2009-07-28 2011-04-19 Tzu-Hsiang Tseng Automatic playing card dispensing system
US8593971B1 (en) 2011-01-25 2013-11-26 Bank Of America Corporation ATM network response diagnostic snapshot
RU2487386C2 (ru) * 2011-03-03 2013-07-10 Федеральное государственное военное образовательное учреждение высшего профессионального образования "Военный авиационный инженерный университет" (г. Воронеж) Министерства обороны Российской Федерации Способ моделирования процессов централизованного управления техническими средствами и система для его осуществления (варианты)
US20120223094A1 (en) * 2011-03-03 2012-09-06 Soft Serve Parts Llc Intelligent monitoring and control system for dispensed chilled food product devices
US20120312049A1 (en) * 2011-03-03 2012-12-13 Soft Serve Parts Llc Intelligent monitoring and control system for dispensed chilled food product devices
US9081656B2 (en) 2011-12-20 2015-07-14 Ncr Corporation Methods and systems for predicting a fault
US9183518B2 (en) * 2011-12-20 2015-11-10 Ncr Corporation Methods and systems for scheduling a predicted fault service call
US8746551B2 (en) * 2012-02-14 2014-06-10 Bank Of America Corporation Predictive fault resolution
WO2014141305A2 (fr) * 2013-03-14 2014-09-18 Narayanakumar Ramanathan Système de gestion de guichet automatique bancaire (atm) pour une résolution instantanée de plaintes de client
US9860109B2 (en) * 2014-05-07 2018-01-02 Getgo, Inc. Automatic alert generation
EP3018638A1 (fr) * 2014-11-07 2016-05-11 Scheidt & Bachmann GmbH Procédé de reconnaissance d'états de fonctionnement exceptionnels dans un dispositif en libre-service
US10643192B2 (en) * 2016-09-06 2020-05-05 Bank Of American Corporation Data transfer between self-service device and server over session or connection in response to capturing sensor data at self-service device
US11182761B2 (en) * 2017-03-03 2021-11-23 Walmart Apollo, Llc Information technology equipment replacement calculation systems and methods
US10643187B2 (en) * 2017-06-09 2020-05-05 Kidde Technologies, Inc. Reporting and prioritizing faults for aircraft downtime reduction
US11113138B2 (en) 2018-01-02 2021-09-07 Carrier Corporation System and method for analyzing and responding to errors within a log file
US10990492B2 (en) * 2018-03-28 2021-04-27 Ncr Corporation Automated terminal problem identification and resolution
JP7360972B2 (ja) 2020-02-25 2023-10-13 東芝テック株式会社 販売データ処理装置およびプログラム
US11176785B1 (en) 2020-06-15 2021-11-16 Bank Of America Corporation Detection of dispensing errors in automated teller machines

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4931963A (en) * 1985-11-22 1990-06-05 Oki Electric Industry Co., Ltd. Self-diagnosis method for a bank note depositing and dispensing machine
US5253184A (en) * 1991-06-19 1993-10-12 Storage Technology Corporation Failure and performance tracking system
US5719383A (en) * 1994-12-29 1998-02-17 Ncr Corporation Transaction terminal and method of maintaining acceptable operation of the transaction terminal
EP0851394A1 (fr) * 1996-12-18 1998-07-01 NCR International, Inc. Machine bancaire
US6098032A (en) * 1996-04-23 2000-08-01 Ascom Hasler Mailing Systems, Inc. System for providing early warning preemptive postal equipment replacement
WO2002018879A1 (fr) * 2000-08-25 2002-03-07 Battelle Memorial Institute Procede et appareil de prediction de la duree de vie restante d'un systeme operationnel
WO2002054223A1 (fr) * 2001-01-05 2002-07-11 Zero Maintenance International Corporation Procede et systeme de prediction de pannes de machines ou de processus

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE69028872T2 (de) 1989-05-18 1997-02-20 Ford Werke Ag Verfahren und Einrichtung zur Diagnose des elektronischen Steuersystems eines Kraftfahrzeuges mit Hilfe der Mustererkennung
US6122575A (en) 1999-06-30 2000-09-19 Hamilton Sundstrand Corporation APU troubleshooting system
DE19959526A1 (de) 1999-12-09 2001-06-13 Bosch Gmbh Robert Verfahren zum Erkennen von Fehlern eines Kraftfahrzeuges
WO2003019377A2 (fr) 2001-08-21 2003-03-06 Idtect Systeme et procede de diagnostic multiniveau echelonnable a distance et maintenance conditionnelle
US7232063B2 (en) * 2003-06-09 2007-06-19 Fujitsu Transaction Solutions Inc. System and method for monitoring and diagnosis of point of sale devices having intelligent hardware

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4931963A (en) * 1985-11-22 1990-06-05 Oki Electric Industry Co., Ltd. Self-diagnosis method for a bank note depositing and dispensing machine
US5253184A (en) * 1991-06-19 1993-10-12 Storage Technology Corporation Failure and performance tracking system
US5719383A (en) * 1994-12-29 1998-02-17 Ncr Corporation Transaction terminal and method of maintaining acceptable operation of the transaction terminal
US6098032A (en) * 1996-04-23 2000-08-01 Ascom Hasler Mailing Systems, Inc. System for providing early warning preemptive postal equipment replacement
EP0851394A1 (fr) * 1996-12-18 1998-07-01 NCR International, Inc. Machine bancaire
WO2002018879A1 (fr) * 2000-08-25 2002-03-07 Battelle Memorial Institute Procede et appareil de prediction de la duree de vie restante d'un systeme operationnel
WO2002054223A1 (fr) * 2001-01-05 2002-07-11 Zero Maintenance International Corporation Procede et systeme de prediction de pannes de machines ou de processus

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
DASGUPTA D ET AL: "Artificial immune system (AIS) research in the last five years", EVOLUTIONARY COMPUTATION, 2003. CEC '03. THE 2003 CONGRESS ON CANBERRA, AUSTRALIA DEC. 8-12, 2003, PISCATAWAY, NJ, USA,IEEE, vol. 1, 8 December 2003 (2003-12-08), pages 123 - 130, XP010706578, ISBN: 0-7803-7804-0 *
SAYEE SUMATHI ET AL: "Self-organized neural network scheme as a data mining tool", PROCEEDINGS OPTICAL DIAGNOSTICS OF LIVING CELLS II, vol. 4057, 6 April 2000 (2000-04-06), US, pages 420 - 427, XP055449607, ISSN: 0277-786X, ISBN: 978-1-5106-1723-0, DOI: 10.1117/12.381761 *

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9355532B2 (en) 2002-11-25 2016-05-31 Diebold Self-Service Systems Division Of Diebold, Incorporated Automated banking machine that is operable to automatically detect and store service activities
US9368003B2 (en) 2002-11-25 2016-06-14 Diebold Self-Service Systems Division Of Diebold, Incorporated Automated banking machine that is operable responsive to data bearing records
US7774649B2 (en) 2008-01-31 2010-08-10 Ncr Corporation Self-service terminal
EP2088564A1 (fr) * 2008-01-31 2009-08-12 NCR Corporation Terminal en libre-service
US8078912B2 (en) 2008-01-31 2011-12-13 Ncr Corporation Self-service terminal
EP2342698A2 (fr) * 2008-09-11 2011-07-13 Non Linear Concepts, Inc. Diagnostic de système intelligent pour machines de traitement de billets de banque
AU2009291573B2 (en) * 2008-09-11 2015-10-15 Toshiba International Corporation Intelligent system diagnostics for bank note processing machines
EP2342698A4 (fr) * 2008-09-11 2012-04-04 Toshiba Int Corp Diagnostic de système intelligent pour machines de traitement de billets de banque
DE102009032023A1 (de) * 2009-07-07 2011-01-13 Giesecke & Devrient Gmbh Banknotenbearbeitungsmaschine mit einer Kassette
WO2014186472A1 (fr) * 2013-05-14 2014-11-20 Diebold Self-Service Systems Division Of Diebold, Incorporated Guichet automatique qui peut fonctionner pour détecter et stocker automatiquement des activités de services
KR20210063397A (ko) * 2018-11-27 2021-06-01 후지츠 프론테크 가부시키가이샤 원인 추정 장치, 원인 추정 출력 방법 및 지엽류 취급 시스템
EP3889864A4 (fr) * 2018-11-27 2021-12-01 Fujitsu Frontech Limited Dispositif d'estimation de cause, procédé de sortie d'estimation de cause, et dispositif de manipulation de feuille de papier
KR102653802B1 (ko) 2018-11-27 2024-04-01 후지츠 프론테크 가부시키가이샤 원인 추정 장치, 원인 추정 출력 방법 및 지엽류 취급 시스템

Also Published As

Publication number Publication date
GB0427694D0 (en) 2005-01-19
US7600671B2 (en) 2009-10-13
US20060131380A1 (en) 2006-06-22

Similar Documents

Publication Publication Date Title
US7600671B2 (en) Method of determining the cause of an error state in an apparatus
US7815103B2 (en) Method of and system for prediction of the state of health of an apparatus
EP0720132B1 (fr) Terminal de transactions
CA2816469C (fr) Prediction de problemes de performance et de defaillances
US9084937B2 (en) Faults and performance issue prediction
EP0851394B1 (fr) Machine bancaire
US11263876B2 (en) Self-service terminal (SST) maintenance and support processing
US9111409B2 (en) Financial transactions processing system including cash automation machine
CN109726090A (zh) 计算系统中性能影响缺陷的识别
WO2012047529A1 (fr) Télémaintenance adaptative de matériel roulant
EP2950281B1 (fr) Procédé et dispositif de détection et commande de qualité de distribution d'argent de terminal en libre service
US20140074708A1 (en) Forecasting of Deposits for a Money Handling Machine
JP5232351B2 (ja) 通貨受容装置
US20140074530A1 (en) Forecasting of Deposit Faults for a Money Handling Machine
JP7404095B2 (ja) 貨幣処理システム、貨幣処理装置、貨幣処理方法及び貨幣処理プログラム
US11556933B2 (en) Money handling system, money handling apparatus, center apparatus, and fraud detecting method for money handling apparatus
CN110619705B (zh) 机芯藏钞检测方法、系统、存储介质及智能设备
JP2001175936A (ja) 自動販売機制御装置、自動販売機および自動販売機管理装置
US20220012973A1 (en) Transaction device
JP5168132B2 (ja) 品質管理システム
JP3132200B2 (ja) 紙幣識別装置
US20230418279A1 (en) Predictive maintenance for terminals
US11907045B2 (en) Information extraction from maintenance logs
US20240070482A1 (en) Terminal power supply degradation detection
US20240346843A1 (en) Counterfeit detection using image analysis

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LI LT LU LV MC NL PL PT RO SE SI SK TR

AX Request for extension of the european patent

Extension state: AL BA HR MK YU

17P Request for examination filed

Effective date: 20061221

17Q First examination report despatched

Effective date: 20070129

AKX Designation fees paid

Designated state(s): DE ES FR GB IT

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: NCR INTERNATIONAL, INC.

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: NCR INTERNATIONAL, INC.

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION HAS BEEN WITHDRAWN

18W Application withdrawn

Effective date: 20180824